legal-small-NER

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

legal-small-NER

This model is a fine-tuned version of nlpaueb/legal-bert-small-uncased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1923

• Accuracy: 0.9546

• Precision: 0.7409

• Recall: 0.7826

• F1: 0.7612

• Classification Report: precision recall f1-score support

     LOC       0.84      0.86      0.85      1668
    MISC       0.54      0.62      0.58       702
     ORG       0.64      0.67      0.65      1661
     PER       0.84      0.89      0.86      1617
  

  micro avg 0.74 0.78 0.76 5648 macro avg 0.71 0.76 0.74 5648 weighted avg 0.74 0.78 0.76 5648

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 2e-05
• train_batch_size: 16
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 3

Training results

     LOC       0.81      0.87      0.84      1668
    MISC       0.52      0.64      0.58       702
     ORG       0.65      0.58      0.61      1661
     PER       0.82      0.89      0.85      1617

micro avg 0.73 0.76 0.75 5648 macro avg 0.70 0.74 0.72 5648 weighted avg 0.73 0.76 0.74 5648 | | 0.0838 | 2.0 | 434 | 0.1941 | 0.9533 | 0.7329 | 0.7817 | 0.7565 | precision recall f1-score support

     LOC       0.84      0.86      0.85      1668
    MISC       0.53      0.59      0.56       702
     ORG       0.62      0.68      0.65      1661
     PER       0.84      0.88      0.86      1617

micro avg 0.73 0.78 0.76 5648 macro avg 0.71 0.76 0.73 5648 weighted avg 0.74 0.78 0.76 5648 | | 0.0447 | 3.0 | 651 | 0.1923 | 0.9546 | 0.7409 | 0.7826 | 0.7612 | precision recall f1-score support

     LOC       0.84      0.86      0.85      1668
    MISC       0.54      0.62      0.58       702
     ORG       0.64      0.67      0.65      1661
     PER       0.84      0.89      0.86      1617

micro avg 0.74 0.78 0.76 5648 macro avg 0.71 0.76 0.74 5648 weighted avg 0.74 0.78 0.76 5648 |

Framework versions

• Transformers 4.30.2
• Pytorch 2.0.0
• Datasets 2.1.0
• Tokenizers 0.13.3